IPB

Welcome Guest ( Log In | Register )

9 Pages V  « < 2 3 4 5 6 > »   
Reply to this topicStart new topic
Cassini image database & updates, And notifications of PDS data releases
ugordan
post Aug 1 2007, 03:09 PM
Post #46


Senior Member
****

Group: Members
Posts: 3648
Joined: 1-October 05
From: Croatia
Member No.: 523



I've stumbled upon an error in the latest batch, there are two polarized UV NAC frames, N1536539385_1.IMG and N1536539427_1.IMG incorrectly labeled as:

TARGET_DESC = "Tethys"
TARGET_LIST = "N/A"
TARGET_NAME = "TETHYS"
OBSERVATION_ID = "ISS_028TE_TETHYSORS001_CIRS"

When in fact they target Enceladus:

Attached Image


IIRC, I've also stumbled upon this incorrectly labeled sequence when browsing VIMS cubes. There in fact was a Tethys ORS sequence running just at about that time - 2006-09-09 23:51:20.0, could this be a case of a quick-n-dirty retargeting to get a couple of Enceladus shots without a dedicated observation?


--------------------
Go to the top of the page
 
+Quote Post
elakdawalla
post Aug 1 2007, 04:03 PM
Post #47


Administrator
****

Group: Admin
Posts: 5172
Joined: 4-August 05
From: Pasadena, CA, USA, Earth
Member No.: 454



I've reduced the thumbnails to 128 pix and converted them from 16-bit PNG to JPG, and together that brings the page down to 2 MB of data, which is much better!

One thing that bothers me about this page is that there's a lot of redundancy -- it doesn't seem necessary to show thumbnails of all of the images of the same spot taken through different filters. I'm going to do some fiddling and see if I can come up with a different way to present this that has less redundancy, but shows more of the metadata.

Hmmm...

--Emily


--------------------
My website - My Patreon - @elakdawalla on Twitter - Please support unmannedspaceflight.com by donating here.
Go to the top of the page
 
+Quote Post
pat
post Aug 2 2007, 02:59 PM
Post #48


Junior Member
**

Group: Members
Posts: 76
Joined: 19-October 05
Member No.: 532



QUOTE (ugordan @ Aug 1 2007, 04:09 PM) *
I've stumbled upon an error in the latest batch, there are two polarized UV NAC frames, N1536539385_1.IMG and N1536539427_1.IMG incorrectly labeled as:

TARGET_DESC = "Tethys"
TARGET_LIST = "N/A"
TARGET_NAME = "TETHYS"
OBSERVATION_ID = "ISS_028TE_TETHYSORS001_CIRS"

When in fact they target Enceladus:

Attached Image


IIRC, I've also stumbled upon this incorrectly labeled sequence when browsing VIMS cubes. There in fact was a Tethys ORS sequence running just at about that time - 2006-09-09 23:51:20.0, could this be a case of a quick-n-dirty retargeting to get a couple of Enceladus shots without a dedicated observation?


Yep, thats down to an error by the person who built the camera command for this observation. The observation, ISS_028TE_TETHYSORS001_CIRS, starts off by turning to Tethys and then tracking Tethys. At the end of the observation Cassini turns from Tethys to Enceladus and ends pointing at Enceladus ready for the next observation to take over. No observations of Enceladus were planned. Margin has to be added to turns to allow for uncertainties so Cassini probably spent 2 mins or so actually pointing at Enceladus right at the end of the observation.

An error was clearly made in syncing the camera command with the spacecraft pointing. When that particular camera command was excecuted the spacecraft was tracking Tethys, thats why 'TETHYS' is the value of the TARGET_NAME keyword. 'Tethys' is the value of the TARGET_DESC keyword becasue thats the photometric model that was used to calculate the exposures for the camera command. The camera command ran 'too long' and images of empty space were taken while Cassini was turning from Tethys to Enceladus, the 12 images from N1536538981_1.IMG to N1536539343_1.IMG. The final two images in the observation were shuttered in that final 2 mins or so of turn margin time when Cassini, or at least the ORS instruments, were pointed at Enceladus. Interestingly the first image taken during the slew, N1536538981_1.IMG is not of empty space but Tethys -- however Tethys is smeared indicating that Cassini must have started turning during the 1.2 seconds that the shutter was open while this image was being taken.

Actually on reflection this can also be explained by the CIRS team changing the spacecraft pointing design after the ISS (and VIMS) Teams had built the camera commands for their rider observations, possibily because the observation had to be shortened.
Go to the top of the page
 
+Quote Post
ugordan
post Aug 2 2007, 03:18 PM
Post #49


Senior Member
****

Group: Members
Posts: 3648
Joined: 1-October 05
From: Croatia
Member No.: 523



Ahh, that explains it. Thanks a lot, pat, I'm always interested in learning in more detail how all these sequences are actually developed and executed at the low level. The shortening of the observation does seem like the most plausible reason. I did notice the one smeared image you mention, but the empty ones must have slipped me as "non-interesting". smile.gif

Speaking of slew maneuvers, one thing I was always wondering is how much buffer time the cameras use when making mosaics of targets spanning over 1 FOV? Is it a couple of minutes or shorter? How fast can Cassini actually turn around say 180 degrees and steady up at the target pointing?


--------------------
Go to the top of the page
 
+Quote Post
elakdawalla
post Aug 2 2007, 05:10 PM
Post #50


Administrator
****

Group: Admin
Posts: 5172
Joined: 4-August 05
From: Pasadena, CA, USA, Earth
Member No.: 454



OK, I've had a slightly different go at it.

http://planetary.org/data/cassini/dioneattempt/index2.html
This is a much improved 726k to download. Note that the links to full-size images still do NOT work.

I used the 1 pix/degree cutoff this time (923 images) and incorporated more metadata, but am only showing one image from each multispectral set. I think I am happy with the number of observations included now -- I think it's more like 44 -- but am not totally happy with the page layout yet. I was hoping to come up with something where I could almost fully automate the generation of the html, but the step of deleting redundant images (something that's really necessary to reduce the amount of data on the browse page) has to be done by hand, and that's time consuming.

Two things I forgot and will include in any future version: I wanted to put the file names in with the metadata, but they're not there (though they are preserved in the URLs, so at least you can see them that way); and I wanted to include the TARGET_LIST field at the end.

EDIT: I added one more thing -- I was curious to see how many of these observations had resulted in products released to Photojournal. Many have, but many have not; and there are also a great many Photojournal releases of images where Dione was smaller than my size cutoff (though many of these were multiple-body images, so the small size made sense). Notes on Photojournal releases are now on the page. Also, the look through Photojournal made me realize that I missed all the opnavs, because these have a target of "SKY" -- I need to modify my query to add in those opnavs. There were 20 or so that were within the size range.

I'm going to start uploading the full-size images now and leave it running overnight -- it's 800 Megs of data!! That's a bit more than I bargained for. Too many 16-bit images.

Comments?

--Emily


--------------------
My website - My Patreon - @elakdawalla on Twitter - Please support unmannedspaceflight.com by donating here.
Go to the top of the page
 
+Quote Post
pat
post Aug 3 2007, 02:02 PM
Post #51


Junior Member
**

Group: Members
Posts: 76
Joined: 19-October 05
Member No.: 532



QUOTE (ugordan @ Aug 2 2007, 04:18 PM) *
Speaking of slew maneuvers, one thing I was always wondering is how much buffer time the cameras use when making mosaics of targets spanning over 1 FOV? Is it a couple of minutes or shorter? How fast can Cassini actually turn around say 180 degrees and steady up at the target pointing?


I'm not sure about the margin for turns while mosaicing but for a 180 degree slew the times are

Roll about X-axis ~30.5 minutes
Roll about Y-axis ~25.8 minutes
Roll about Z-axis ~16.4 minutes

using high power for the reaction wheels
and

Roll about X-axis ~58 minutes
Roll about Y-axis ~42 minutes
Roll about Z-axis ~38 minutes

using low power for the reaction wheels

and a 180 degree RCS turn is about 11 minutes

The settling time after a turn is nominally about 60 seconds I think.
Go to the top of the page
 
+Quote Post
ugordan
post Aug 3 2007, 02:32 PM
Post #52


Senior Member
****

Group: Members
Posts: 3648
Joined: 1-October 05
From: Croatia
Member No.: 523



Wow, I assumed the reaction wheel figures were high but these are quite substantial indeed. Is the time difference primarily due to different mass distribution along each axis? I can't seem to dig up a schematic showing the orientation of principal axes, but this would suggest the Z axis running along the length of the s/c, parallel with HGA boresight and the X axis being the most inert of them, orthogonal to the plane in which HGA boresight and magnetometer boom lie?

How do these figures change with propellant depletion (if at all)? Optimizing slew maneuvers is no doubt one of the (many) factors in increasing useful observation times...


--------------------
Go to the top of the page
 
+Quote Post
ugordan
post Aug 3 2007, 02:45 PM
Post #53


Senior Member
****

Group: Members
Posts: 3648
Joined: 1-October 05
From: Croatia
Member No.: 523



QUOTE (elakdawalla @ Aug 2 2007, 06:10 PM) *
I'm going to start uploading the full-size images now and leave it running overnight -- it's 800 Megs of data!! That's a bit more than I bargained for. Too many 16-bit images.

Comments?

I really like the new layout, much more bandwidth friendly and the color-coded filters are also a nice touch. How did you find all those Photojournal releases derived from Dione sequences - manually by dates?

One thing regarding the full size PNGs. They are 16 bit, but are very dark, perhaps a uniform scaling factor could be applied to all images to maximize DN in the PNGs. IIRC Bjorn's IMG2PNG does implement this feature. This would ideally be the highest scaling factor that keeps the brightest albedo object in the Saturnian system (i.e. Enceladus) from being overexposed, of course taking into account its weird zero-phase geometric albedo of 1.375. I don't know what scaling factor you currently use, but if low enough it might in fact reduce s/n ratio if the 12bit images are fit into 8 bit values.


--------------------
Go to the top of the page
 
+Quote Post
elakdawalla
post Aug 3 2007, 04:47 PM
Post #54


Administrator
****

Group: Admin
Posts: 5172
Joined: 4-August 05
From: Pasadena, CA, USA, Earth
Member No.: 454



Thanks. For the Photojournal images, I just searched on Dione, and then I matched the dates and, when date was ambiguous (or, in one case, incorrect), I matched the phase angles and distances. A bit tedious but that's one thing I've been very curious about -- just how many of the "good" images have been released to the public in one way or another? It's not a bad proportion, it turns out, but it's not all of them by any means, and there is a LOT that can be done that hasn't yet been done with color data.

The 16-bit images do have a variety of brightnesses, but I did use IMG2PNG with the "-r" switch and all the Cassini calibration data to calibrate them as I did the format conversion. Clear and green filter images do seem to be pretty reasonable in their brightness, UV ones are pretty dark -- which is more or less what I expect, actually. Remember that Dione is somewhat dark as icy satellites go. Maybe Bjorn can comment on whether the full-size images look like they are supposed to, especially in the UV. I did get errors on every WAC image upon conversion though, so I'm not sure that those are properly calibrated. Of the 900 or so NAC images, IMG2PNG reported errors during the calibration step for about 20 of them.

On the next go-round I'll do Enceladus and we can see how different those look.

--Emily


--------------------
My website - My Patreon - @elakdawalla on Twitter - Please support unmannedspaceflight.com by donating here.
Go to the top of the page
 
+Quote Post
ugordan
post Aug 3 2007, 06:36 PM
Post #55


Senior Member
****

Group: Members
Posts: 3648
Joined: 1-October 05
From: Croatia
Member No.: 523



I was intrigued by some fairly big brightness differences between different filters in Emily's page. For example the UV images turn out much darker than when I throw them through my software. I decided to do a comparison of narrow angle frames N1507738546_2 (RED,CL2), N1507738663_2 (CL1,GRN) and N1507738491_2 (BL1,CL2). Below are simple RGB composites, modified only to resize and register the three filters:


The left image uses IMG2PNG converted data, the right one is my calibration result.

I didn't touch the channel balance in the left case, there are a couple of notes about my results:
I do all the calibration steps except dark current removal. In practice this doesn't amount to anything than a slightly brighter background signal in the longer exposed images (blue and especially UV in NAC and violet in WAC case) and is in no way capable of such a drastic difference. If you really kick up the saturation in my image, you'll see the bluish hue likely due to dark current.
Other than that, I use the official calibration "fudge factors" found on COISS_0011 with a slight modification for RED/CL2 filters. Early on I realized the red channel turned out consistently too bright with those factors so I worked out it needed to be multplied by 0.949 to bring it in line with other filters. The official calibration factors might indeed be off since they're quite old as a contact from PDS emailed me, in fact the whole calibration procedure was greatly improved since then. I don't have too much faith in other infrared filter factors either. Such factors weren't even derived for WAC and are missing from the calibration volume.

The color of Dione's surface in the left case suggests to me it shows the total radiance as seen by the instrument. It reminds me of my VIMS experiments without dividing by solar spectrum as sunlight is actually yellowish.
In my calibration procedure the filter transmissivities are divided by integrated solar spectrum and produce reflectance (I hope it's the right term) instead, the intrinsic color of the moon if you will. This is consistent in that it produces gray Mimas, Enceladus and gray non-stained hemispheres of Dione and Tethys. In essence the color they'd appear if shined upon by a big, white light lamp.


--------------------
Go to the top of the page
 
+Quote Post
MouseOnMars
post Aug 3 2007, 07:30 PM
Post #56


Junior Member
**

Group: Members
Posts: 76
Joined: 4-June 07
From: United Kingdom
Member No.: 2288



I like your page elakdawalla smile.gif

Reminds me of a similiar thing I did for Themis. Apologies for the file share download, but it's all I have at the moment.

MouseOnMars


--------------------
Go to the top of the page
 
+Quote Post
ugordan
post Sep 30 2007, 03:03 PM
Post #57


Senior Member
****

Group: Members
Posts: 3648
Joined: 1-October 05
From: Croatia
Member No.: 523



It's starting to appear, the ISS volumes COISS_2026 and COISS_2027 are up, but not searchable yet. They can be accessed directly:
http://pdsimg.jpl.nasa.gov/data/cassini/ca...ter/coiss_2026/
http://pdsimg.jpl.nasa.gov/data/cassini/ca...ter/coiss_2027/

I've done a preliminary browse of the thumbnails to see if anything interesting's up. This is still the time period when Cassini was spending a lot of time on the nightside, seeing unlit rings. The real good Saturn stuff should be appearing in the Jan 1, 2008 release.

In the meantime, here are a few takeouts:

Hyperion in natural color, 2x magnified.


Iapetus in natural and IR3/GRN/UV3 color, 2x magnified. November 27, 2006.


Couple of Saturn north pole shots. A moon shadow (probably Mimas) visible near the terminator in the left image.

Titan and a lovely shot with Saturn's hazy limb showing through the translucent, unlit rings:


--------------------
Go to the top of the page
 
+Quote Post
elakdawalla
post Oct 1 2007, 03:58 PM
Post #58


Administrator
****

Group: Admin
Posts: 5172
Joined: 4-August 05
From: Pasadena, CA, USA, Earth
Member No.: 454



I've been waiting for this!! Thanks for the notice, Gordan.

Bjorn, I'm ready to make the database update whenever you can put the tables together.

--Emily


--------------------
My website - My Patreon - @elakdawalla on Twitter - Please support unmannedspaceflight.com by donating here.
Go to the top of the page
 
+Quote Post
volcanopele
post Oct 1 2007, 05:09 PM
Post #59


Senior Member
****

Group: Moderator
Posts: 3225
Joined: 11-February 04
From: Tucson, AZ
Member No.: 23



This also means that some of my favorite SAR swaths, from T19, T20 (HiSAR over Tortola Facula), and T21 will also be released.


--------------------
&@^^!% Jim! I'm a geologist, not a physicist!
The Gish Bar Times - A Blog all about Jupiter's Moon Io
Go to the top of the page
 
+Quote Post
Bjorn Jonsson
post Oct 10 2007, 12:08 AM
Post #60


IMG to PNG GOD
****

Group: Moderator
Posts: 2250
Joined: 19-February 04
From: Near fire and ice
Member No.: 38



The information from this ISS release is now ready for getting added to the Access database described in this thread (see this message). The database includes all of the information from the index.tab files and makes the images much easier to work with.

As before, Emily has hosted them on the same server as the database:

http://filicio.us/tpss3/files/31020/cassin...es_index_26.zip
http://filicio.us/tpss3/files/31032/cassin...es_index_27.zip

These are Excel files that must be added to the Access database as described here.
Go to the top of the page
 
+Quote Post

9 Pages V  « < 2 3 4 5 6 > » 
Reply to this topicStart new topic

 



RSS Lo-Fi Version Time is now: 28th March 2024 - 11:27 AM
RULES AND GUIDELINES
Please read the Forum Rules and Guidelines before posting.

IMAGE COPYRIGHT
Images posted on UnmannedSpaceflight.com may be copyrighted. Do not reproduce without permission. Read here for further information on space images and copyright.

OPINIONS AND MODERATION
Opinions expressed on UnmannedSpaceflight.com are those of the individual posters and do not necessarily reflect the opinions of UnmannedSpaceflight.com or The Planetary Society. The all-volunteer UnmannedSpaceflight.com moderation team is wholly independent of The Planetary Society. The Planetary Society has no influence over decisions made by the UnmannedSpaceflight.com moderators.
SUPPORT THE FORUM
Unmannedspaceflight.com is funded by the Planetary Society. Please consider supporting our work and many other projects by donating to the Society or becoming a member.